Voices and caption styles
The narration and captions in a faceless video carry most of the personality. Schedulin gives you full control over both.
Built-in voices
Schedulin ships with six ElevenLabs voices, each with a preview button in the editor:
- Rachel — calm, clear American English. The default; works for most niches.
- Sarah — warm, professional female. Great for fact and edutainment channels.
- Domi — young, upbeat female. Suits social-media-native, conversational scripts.
- Antoni — well-rounded male, American. Versatile for storytelling.
- Arnold — crisp, confident male. Works for finance and authority niches.
- Adam — deep, authoritative male. Default pick for motivation and personal finance.
Click any voice to preview before generating.
Custom voice cloning
You can use any voice you've cloned through ElevenLabs by pasting the voice ID into the custom voice field. Many of our power users clone their own voice once and reuse it across hundreds of faceless videos — giving the channel a consistent on-brand sound without ever recording a take.
To clone a voice, see ElevenLabs' Voice Lab docs. Once cloned, copy the voice ID from your ElevenLabs dashboard and paste it into Schedulin.
Caption styles
All three styles are karaoke-timed — words highlight in sync with the audio using word-level timestamps from ElevenLabs.
Bold yellow
Large bold sans-serif with a yellow highlight on the active word. The classic look for motivation, finance, and high-energy edutainment shorts. Maximum readability, maximum 'TikTok'.
TikTok white
Clean white text with a soft drop shadow. The active word grows slightly. Suits storytelling, lifestyle, and any niche where bold yellow would feel too aggressive.
Minimal
Small, lowercase, no highlight — the active word fades in subtly. Best for cinematic AI-stills content (history, scary stories) where you don't want captions to dominate the image.
Tips
- Caption style is set per render — you can re-render the same script with a different style without rewriting anything.
- The captions always use the exact text of your script. Edit punctuation in the script field (not after generation) if you want timing changes.